CLAVATELLA PROLIFERA. 215 



sion from the canal system prolongs itself into their axes. They are now plainly seen to be the 

 rudimental marginal tentacles, as yet, however, quite simple. 



We next observe a small tubercle showing itself close to the extremity and upon the inner 

 side of each of the developing marginal tentacles (fig. 10), so that the tentacles at this stage appear 

 obscurely double-headed, the outer head set with thread-cells, the inner one without any special 

 accumulation of these bodies. The two heads become rapidly more and more distinct from 

 one another, each borne on a stalk, and the tentacle thus becomes bifurcated at its extremity. 

 The thread-cells increase in number on the extremity of the onter branch, which now terminates 

 in a distinct capitulum, while a suctorial disc of attachment is developed on the extremity of the 

 inner branch. The ocellus is by this time quite distinct, and we soon find that the little medusa, 

 fitted for an independent existence, has detached itself from its stalk and creeps about on the 

 corallines and sea-weeds of its rock-pool, the inner branch of its tentacles serving for attachment 

 and locomotion, the other aiding it apparently in the prehension of its prey (fig. 4). 



It will be noticed that in the development just traced an ectotheca is never present, the 

 medusa being exposed during the whole time to the direct contact of the surrounding water. 



The important observation was made by Krohn that the C/avatelia-medusa may repeat itself 

 by buds from the medusa itself, and he has further shown that this gemmation may take place 

 simultaneously with sexual reproduction, the buds being borne by medusae which are at the same 

 time laden with ova. 



I can fully confirm these observations of Krohn. The buds are tlevcloped from the margin 

 of the disc, and from the middle of the interval between two tentacles (fig. 5, a, b). They arc 

 almost always two in number, and are then pi-oduced in two opposite interradial spaces, from 

 which they project outwards between the roots of the tentacles. They originate in the form of a 

 small hernia-like process from the circular canal, consisting of a layer of ectoderm lined by one 

 of endodcrm and having its cavity freely communicating with that of the circular canal. Their 

 development is entirely similar to that of the medusas budded from the hydranth. 



I found the buds produced both by medusae in which eggs were at the same time present 

 and by those in which eggs had not yet shown themselves. 



A very remarkable difference from the position just described as that of the buds in 

 Clavatella proUfera had, however, been noticed by Claparede in the species which came under his 

 observation. In this the buds, while still produced from the circular canal in the interradial 

 spaces, are described as developing themselves under the ectoderm, so as to lie between the disc of 

 the medusa and its digestive cavity. It is difficult to say where the space exists for them in this 

 direction unless that in their growth they insinuate themselves between the ectoderm of the disc 

 and the endoderm of the enlarged atrium. Indeed, Claparede expresses some doubt as to their 

 being really buds, and thinks it possible that they may be eggs which had passed on to an 

 advanced stage of development in the proper egg-cavity, which in this species is described as 

 lying beneath the ectoderm of the lower side. 



A comparison of the medusa of Clavatella with an ordinary hydroid medusa gives results of 

 much interest. 



It will be borne in mind that in many medusae, such, for example, as those of Synconjne 

 and Bougainvillia, the proximal end of the manubrium or that from which the radiating canals are 

 given off presents a special dilatation, and has those endodermal cells which are characterised 

 by the secretion in them of the coloui'ed granules more abundantly developed there than elsewhere. 



